Chelsea - ManCity

Match report

from IPython.display import Image, display
display(Image(filename='img/cover.png', embed=True))

Result

display(Image(filename='img/result.png', embed=True))

Standings before the game

display(Image(filename='img/standings.png', embed=True))

Team form

display(Image(filename='img/team_form.png', embed=True))

%pip install pandas

Requirement already satisfied: pandas in /usr/local/lib/python3.11/site-packages (1.5.3)
Requirement already satisfied: python-dateutil>=2.8.1 in /Users/ivn/Library/Python/3.11/lib/python/site-packages (from pandas) (2.8.2)
Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.11/site-packages (from pandas) (2023.3.post1)
Requirement already satisfied: numpy>=1.21.0 in /usr/local/lib/python3.11/site-packages (from pandas) (1.25.0)
Requirement already satisfied: six>=1.5 in /Users/ivn/Library/Python/3.11/lib/python/site-packages (from python-dateutil>=2.8.1->pandas) (1.16.0)

[notice] A new release of pip is available: 23.3.1 -> 23.3.2
[notice] To update, run: python3.11 -m pip install --upgrade pip
Note: you may need to restart the kernel to use updated packages.

%pip install soccerbars

Collecting soccerbars
  Downloading soccerbars-0.2.1-py3-none-any.whl.metadata (10 kB)
Requirement already satisfied: matplotlib<4.0.0,>=3.7.1 in /usr/local/lib/python3.11/site-packages (from soccerbars) (3.8.2)
Requirement already satisfied: numpy<2.0.0,>=1.24.0 in /usr/local/lib/python3.11/site-packages (from soccerbars) (1.25.0)
Requirement already satisfied: contourpy>=1.0.1 in /usr/local/lib/python3.11/site-packages (from matplotlib<4.0.0,>=3.7.1->soccerbars) (1.2.0)
Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.11/site-packages (from matplotlib<4.0.0,>=3.7.1->soccerbars) (0.12.1)
Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.11/site-packages (from matplotlib<4.0.0,>=3.7.1->soccerbars) (4.47.0)
Requirement already satisfied: kiwisolver>=1.3.1 in /usr/local/lib/python3.11/site-packages (from matplotlib<4.0.0,>=3.7.1->soccerbars) (1.4.5)
Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.11/site-packages (from matplotlib<4.0.0,>=3.7.1->soccerbars) (23.2)
Requirement already satisfied: pillow>=8 in /usr/local/lib/python3.11/site-packages (from matplotlib<4.0.0,>=3.7.1->soccerbars) (10.1.0)
Requirement already satisfied: pyparsing>=2.3.1 in /usr/local/lib/python3.11/site-packages (from matplotlib<4.0.0,>=3.7.1->soccerbars) (3.1.1)
Requirement already satisfied: python-dateutil>=2.7 in /Users/ivn/Library/Python/3.11/lib/python/site-packages (from matplotlib<4.0.0,>=3.7.1->soccerbars) (2.8.2)
Requirement already satisfied: six>=1.5 in /Users/ivn/Library/Python/3.11/lib/python/site-packages (from python-dateutil>=2.7->matplotlib<4.0.0,>=3.7.1->soccerbars) (1.16.0)
Downloading soccerbars-0.2.1-py3-none-any.whl (14 kB)
Installing collected packages: soccerbars
Successfully installed soccerbars-0.2.1

[notice] A new release of pip is available: 23.3.1 -> 23.3.2
[notice] To update, run: python3.11 -m pip install --upgrade pip
Note: you may need to restart the kernel to use updated packages.

import pandas as pd
from soccerbars import soccerbar

#df_team forms = pd.read_csv('../data/FBRef/matches_season_2023-2024.csv')

df_team_forms = pd.read_csv('data/FBRef/matches_season_2023-2024.csv')

df_team_forms = df_team_forms.drop(columns=['Link', 'Home Team ID', 'Away Team ID', 'Home xG', 'Away xG', 'Home npxG', 'Away npxG', 'Home xAG', 'Away xAG', 'Shots', 'Home Player Stats', 'Away Player Stats', 'Home Ast', 'Away Ast'])
df_team_forms = df_team_forms.loc[df_team_forms['Date'] < '2023-11-12']
mcity = df_team_forms.loc[df_team_forms['Away Team'].isin(['Manchester City']) | df_team_forms['Home Team'].isin(['Manchester City'])]
mcity

	Unnamed: 0	Date	Stage	Home Team	Away Team	Home Formation	Away Formation	Home Goals	Away Goals
0	0	2023-08-11	Matchweek 1	Burnley	Manchester City	5-4-1	4-2-3-1	0	3
15	15	2023-08-19	Matchweek 2	Manchester City	Newcastle United	4-2-3-1	4-3-3	1	0
26	26	2023-08-27	Matchweek 3	Sheffield United	Manchester City	3-5-2	4-2-3-1	1	2
35	35	2023-09-02	Matchweek 4	Manchester City	Fulham	4-2-3-1	4-3-3	5	1
42	42	2023-09-16	Matchweek 5	West Ham United	Manchester City	4-2-3-1	4-2-3-1	1	3
49	49	2023-09-23	Matchweek 6	Manchester City	Nottingham Forest	4-2-3-1	5-4-1	2	0
63	63	2023-09-30	Matchweek 7	Wolverhampton Wanderers	Manchester City	3-4-3	4-2-3-1	2	1
79	79	2023-10-08	Matchweek 8	Arsenal	Manchester City	4-3-3	4-3-3	1	0
82	82	2023-10-21	Matchweek 9	Manchester City	Brighton & Hove Albion	3-2-4-1	4-2-3-1	2	1
97	97	2023-10-29	Matchweek 10	Manchester United	Manchester City	4-2-3-1	4-2-3-1	0	3
104	104	2023-11-04	Matchweek 11	Manchester City	Bournemouth	3-4-3	5-4-1	6	1

chelsea = df_team_forms.loc[df_team_forms['Away Team'].isin(['Chelsea']) | df_team_forms['Home Team'].isin(['Chelsea'])]
chelsea

	Unnamed: 0	Date	Stage	Home Team	Away Team	Home Formation	Away Formation	Home Goals	Away Goals
8	8	2023-08-13	Matchweek 1	Chelsea	Liverpool	3-4-3	4-3-3	1	1
17	17	2023-08-20	Matchweek 2	West Ham United	Chelsea	4-2-3-1	3-4-3	3	1
19	19	2023-08-25	Matchweek 3	Chelsea	Luton Town	3-4-3	5-3-2	3	0
33	33	2023-09-02	Matchweek 4	Chelsea	Nottingham Forest	3-4-3	3-4-3	0	1
46	46	2023-09-17	Matchweek 5	Bournemouth	Chelsea	4-2-3-1	4-2-3-1	0	0
56	56	2023-09-24	Matchweek 6	Chelsea	Aston Villa	4-2-3-1	4-2-3-1	0	1
68	68	2023-10-02	Matchweek 7	Fulham	Chelsea	4-3-3	4-3-3	0	2
72	72	2023-10-07	Matchweek 8	Burnley	Chelsea	4-3-3	4-3-3	1	4
81	81	2023-10-21	Matchweek 9	Chelsea	Arsenal	4-2-3-1	4-3-3	2	2
91	91	2023-10-28	Matchweek 10	Chelsea	Brentford	4-2-3-1	3-5-2	0	2
109	109	2023-11-06	Matchweek 11	Tottenham Hotspur	Chelsea	4-2-3-1	4-2-3-1	1	4

ManCity results before the match

isaway = []
home_goals = []
away_goals = []
for index, row in mcity.iterrows():
    if row['Away Team'] == 'Manchester City':
        isaway.append(True)
    else: isaway.append(False)
    home_goals.append(row['Home Goals'])
    away_goals.append(row['Away Goals'])

soccerbar([
    home_goals, 
    away_goals, 
    isaway
], outlined=True)

<Axes: >

Chelsea results before the match

isaway = []
home_goals = []
away_goals = []
for index, row in chelsea.iterrows():
    if row['Away Team'] == 'Chelsea':
        isaway.append(True)
    else: isaway.append(False)
    home_goals.append(row['Home Goals'])
    away_goals.append(row['Away Goals'])

soccerbar([
    home_goals, 
    away_goals, 
    isaway
], outlined=True)

<Axes: >

display(Image(filename='img/head-to-head.png', embed=True))

display(Image(filename='img/historical_head-to-head.png', embed=True))

Line up

display(Image(filename='img/lineup.png', embed=True))

display(Image(filename='img/sabstitutes.png', embed=True))

display(Image(filename='img/lineup_worth.png', embed=True))

Average players positions

from mplsoccer.pitch import Pitch
import pandas as pd

# spadl format assumes pitch size 68 x 105,
# so we use pitch_type=`uefa` for mplsoccer
# https://socceraction.readthedocs.io/en/latest/documentation/spadl/spadl.html

def get_pitch():
  return Pitch(pitch_type='uefa',
              pitch_color='white',
              line_color='#c7d5cc')

import soccerdata as sd
import os
from pathlib import Path

# https://www.whoscored.com/Matches/1729243/Live/England-Premier-League-2023-2024-Chelsea-Manchester-City

data_dir=Path(f"{os.getcwd()}/data/soccerdata_cache")
ws = sd.WhoScored(leagues="ENG-Premier League", seasons="23-24", data_dir=data_dir)

MC_team_id = 167
chelsea_team_id = 15
GAME_ID=1729243
events_atomic_spadl = ws.read_events(match_id=GAME_ID, force_cache=True, output_fmt="atomic-spadl")

[01/30/24 21:19:42] INFO     Saving cached data to                                                    _common.py:91
                             /Users/ivn/Documents/projects/fa_ucu/Manchester-City/data/soccerdata_cac              
                             he                                                                                    

[01/30/24 21:20:01] INFO     patching driver executable /Users/ivn/Library/Application               patcher.py:346
                             Support/undetected_chromedriver/undetected_chromedriver                               

[01/30/24 21:20:04] INFO     Retrieving game schedule of ENG-Premier League - 2324 from the cache  whoscored.py:437

                    INFO     [1/1] Retrieving game with id=1729243                                 whoscored.py:693

events_raw = ws.read_events(match_id=GAME_ID, force_cache=True, output_fmt="raw")
events_spadl = ws.read_events(match_id=GAME_ID, force_cache=True, output_fmt="spadl")
events_atomic_spadl = ws.read_events(match_id=GAME_ID, force_cache=True, output_fmt="atomic-spadl")
events = ws.read_events(match_id=GAME_ID, force_cache=True, output_fmt="events")

[01/30/24 21:21:33] INFO     Retrieving game schedule of ENG-Premier League - 2324 from the cache  whoscored.py:437

                    INFO     [1/1] Retrieving game with id=1729243                                 whoscored.py:693

                    INFO     Retrieving game schedule of ENG-Premier League - 2324 from the cache  whoscored.py:437

                    INFO     [1/1] Retrieving game with id=1729243                                 whoscored.py:693

                    INFO     Retrieving game schedule of ENG-Premier League - 2324 from the cache  whoscored.py:437

                    INFO     [1/1] Retrieving game with id=1729243                                 whoscored.py:693

                    INFO     Retrieving game schedule of ENG-Premier League - 2324 from the cache  whoscored.py:437

                    INFO     [1/1] Retrieving game with id=1729243                                 whoscored.py:693

# Average positions for all players (including subs!)

def avg_loc_by_team(team_id: int):
    avg_loc = events_atomic_spadl[events_atomic_spadl['team_id'] == team_id].groupby('player').agg({'x': 'mean', 'y': 'mean'})

    pitch = get_pitch()

    fig, ax = pitch.draw()

    sc = pitch.scatter(x=avg_loc['x'].to_list(), 
                    y=avg_loc['y'].to_list(),
                    # color=df_shot['color'].to_list(),
                    # marker='football',
                    ax=ax)

    n = avg_loc.index.to_list()
    data = sc.get_offsets()

    # leg = ax.legend(borderpad=1, markerscale=0.5, labelspacing=1.5, loc='upper center', fontsize=15)

    for idx,label in enumerate(n):
        ax.annotate(label, (data[idx][0]+2, data[idx][1]))

MC_team_id = 167
chelsea_team_id = 15

avg_loc_by_team(MC_team_id)
avg_loc_by_team(chelsea_team_id)

Match stats

display(Image(filename='img/stats.png', embed=True))

Shots

%pip install pandas

Requirement already satisfied: pandas in /usr/local/lib/python3.11/site-packages (1.5.3)
Requirement already satisfied: python-dateutil>=2.8.1 in /Users/ivn/Library/Python/3.11/lib/python/site-packages (from pandas) (2.8.2)
Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.11/site-packages (from pandas) (2023.3.post1)
Requirement already satisfied: numpy>=1.21.0 in /usr/local/lib/python3.11/site-packages (from pandas) (1.25.0)
Requirement already satisfied: six>=1.5 in /Users/ivn/Library/Python/3.11/lib/python/site-packages (from python-dateutil>=2.8.1->pandas) (1.16.0)

[notice] A new release of pip is available: 23.3.1 -> 23.3.2
[notice] To update, run: python3.11 -m pip install --upgrade pip
Note: you may need to restart the kernel to use updated packages.

import pandas as pd
import matplotlib.pyplot as plt
from mplsoccer.pitch import Pitch, VerticalPitch
import seaborn as sns

df_1 = pd.read_csv('data/12112023/events_1729243.csv')

# define dataframe with shots and goals
shots = df_1.loc[(df_1['type'] == 'SavedShot') | (df_1['type'] == 'Goal')].set_index('id')
# define team names
team1, team2 = df_1.team.unique()

# check df
shots.tail()

	league	season	game	game_id	period	minute	second	expanded_minute	type	outcome_type	...	goal_mouth_z	blocked_x	blocked_y	qualifiers	is_touch	is_shot	is_goal	card_type	related_event_id	related_player_id
id
2617386379	ENG-Premier League	2324	2023-11-12 Chelsea-Manchester City	1729243	SecondHalf	66	22.0	73	SavedShot	Successful	...	1.3	97.2	50.5	[{'type': {'displayName': 'Assisted', 'value':...	True	True	NaN	NaN	546.0	410175.0
2617386415	ENG-Premier League	2324	2023-11-12 Chelsea-Manchester City	1729243	SecondHalf	66	25.0	73	Goal	Successful	...	1.9	NaN	NaN	[{'type': {'displayName': 'StandingSave', 'val...	True	True	True	NaN	NaN	NaN
2617404351	ENG-Premier League	2324	2023-11-12 Chelsea-Manchester City	1729243	SecondHalf	85	30.0	92	SavedShot	Successful	...	19.0	83.6	49.5	[{'type': {'displayName': 'LowCentre', 'value'...	True	True	NaN	NaN	819.0	315227.0
2617404379	ENG-Premier League	2324	2023-11-12 Chelsea-Manchester City	1729243	SecondHalf	85	33.0	92	Goal	Successful	...	14.6	NaN	NaN	[{'type': {'displayName': 'StandingSave', 'val...	True	True	True	NaN	822.0	315227.0
2617410983	ENG-Premier League	2324	2023-11-12 Chelsea-Manchester City	1729243	SecondHalf	94	24.0	101	Goal	Successful	...	22.2	NaN	NaN	[{'type': {'displayName': 'HighLeft', 'value':...	True	True	True	NaN	NaN	NaN

5 rows × 29 columns

pitch = Pitch(pitch_type='opta', line_color = 'white', pitch_color = 'green')
fig, ax = pitch.draw(figsize=(12,8))

# plot the shots by looping through them
for i,shot in shots.iterrows():
    # get the information
    x=shot['x']
    y=shot['y']
    goal=shot['type']=='Goal'
    team_name=shot['team']
    # set circlesize
    circleSize=1
    # plot Chelsea
    if (team_name==team1):
        if goal:
            shotCircle=plt.Circle((x,y),circleSize,color="blue")
            plt.text(x+1,y-1,shot['player'])
        else:
            shotCircle=plt.Circle((x,y),circleSize,color="blue")
            shotCircle.set_alpha(.2)
    # plot Man City
    else:
        if goal:
            shotCircle=plt.Circle((100-x,100 - y),circleSize,color="red") #https://mplsoccer.readthedocs.io/en/latest/gallery/pitch_setup/plot_compare_pitches.html
            plt.text(100-x+1,100 - y - 1, shot['player']) #https://mplsoccer.readthedocs.io/en/latest/gallery/pitch_setup/plot_compare_pitches.html
        else:
            shotCircle=plt.Circle((100-x,100 - y),circleSize,color="red")
            shotCircle.set_alpha(.2)
    ax.add_patch(shotCircle)
    # set title
fig.suptitle("Chelsea (blue) & ManCity (red) shots",  fontsize=20, x=0.5, y=0.95)
fig.set_size_inches(10, 7)
plt.show()

Raheem Sterling goal

from mplsoccer.pitch import Pitch

# points that led to a goal at event #1015 (Sterling, goal #3)
shot = 1015
df_shot = events_atomic_spadl[shot-4:shot+1].copy()
df_shot['color'] = 'red'
df_shot.loc[shot, 'color'] = 'green'

pitch = get_pitch()
fig, ax = pitch.draw()

print(df_shot.loc[shot])
_ = pitch.scatter(x=df_shot['x'].to_list(), 
                  y=df_shot['y'].to_list(),
                  color=df_shot['color'].to_list(),
                  ax=ax)

game_id                      1729243
original_event_id         2617331399
action_id                       1015
period_id                          1
time_seconds                  2183.0
team_id                           15
player_id                    97692.0
x                              105.0
y                             35.088
dx                               0.0
dy                               0.0
type_id                           27
bodypart_id                        5
player               Raheem Sterling
team                         Chelsea
color                          green
Name: 1015, dtype: object

%pip install socceraction

Requirement already satisfied: socceraction in /usr/local/lib/python3.11/site-packages (1.4.2)
Requirement already satisfied: lxml<5.0.0,>=4.6.3 in /usr/local/lib/python3.11/site-packages (from socceraction) (4.9.4)
Requirement already satisfied: numpy<2.0.0,>=1.21.2 in /usr/local/lib/python3.11/site-packages (from socceraction) (1.25.0)
Requirement already satisfied: pandas<2.0.0,>=1.3.3 in /usr/local/lib/python3.11/site-packages (from socceraction) (1.5.3)
Requirement already satisfied: pandera<0.14.0,>=0.13.4 in /usr/local/lib/python3.11/site-packages (from socceraction) (0.13.4)
Requirement already satisfied: scikit-learn<2.0.0,>=1.2.1 in /usr/local/lib/python3.11/site-packages (from socceraction) (1.3.2)
Requirement already satisfied: python-dateutil>=2.8.1 in /Users/ivn/Library/Python/3.11/lib/python/site-packages (from pandas<2.0.0,>=1.3.3->socceraction) (2.8.2)
Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.11/site-packages (from pandas<2.0.0,>=1.3.3->socceraction) (2023.3.post1)
Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.11/site-packages (from pandera<0.14.0,>=0.13.4->socceraction) (23.2)
Requirement already satisfied: pydantic in /usr/local/lib/python3.11/site-packages (from pandera<0.14.0,>=0.13.4->socceraction) (2.5.3)
Requirement already satisfied: typing-inspect>=0.6.0 in /usr/local/lib/python3.11/site-packages (from pandera<0.14.0,>=0.13.4->socceraction) (0.9.0)
Requirement already satisfied: wrapt in /usr/local/lib/python3.11/site-packages (from pandera<0.14.0,>=0.13.4->socceraction) (1.16.0)
Requirement already satisfied: scipy>=1.5.0 in /usr/local/lib/python3.11/site-packages (from scikit-learn<2.0.0,>=1.2.1->socceraction) (1.11.4)
Requirement already satisfied: joblib>=1.1.1 in /usr/local/lib/python3.11/site-packages (from scikit-learn<2.0.0,>=1.2.1->socceraction) (1.3.2)
Requirement already satisfied: threadpoolctl>=2.0.0 in /usr/local/lib/python3.11/site-packages (from scikit-learn<2.0.0,>=1.2.1->socceraction) (3.2.0)
Requirement already satisfied: six>=1.5 in /Users/ivn/Library/Python/3.11/lib/python/site-packages (from python-dateutil>=2.8.1->pandas<2.0.0,>=1.3.3->socceraction) (1.16.0)
Requirement already satisfied: mypy-extensions>=0.3.0 in /usr/local/lib/python3.11/site-packages (from typing-inspect>=0.6.0->pandera<0.14.0,>=0.13.4->socceraction) (1.0.0)
Requirement already satisfied: typing-extensions>=3.7.4 in /usr/local/lib/python3.11/site-packages (from typing-inspect>=0.6.0->pandera<0.14.0,>=0.13.4->socceraction) (4.9.0)
Requirement already satisfied: annotated-types>=0.4.0 in /usr/local/lib/python3.11/site-packages (from pydantic->pandera<0.14.0,>=0.13.4->socceraction) (0.6.0)
Requirement already satisfied: pydantic-core==2.14.6 in /usr/local/lib/python3.11/site-packages (from pydantic->pandera<0.14.0,>=0.13.4->socceraction) (2.14.6)

[notice] A new release of pip is available: 23.3.1 -> 23.3.2
[notice] To update, run: python3.11 -m pip install --upgrade pip
Note: you may need to restart the kernel to use updated packages.

%pip install matplotsoccer

Requirement already satisfied: matplotsoccer in /usr/local/lib/python3.11/site-packages (0.0.8)

[notice] A new release of pip is available: 23.3.1 -> 23.3.2
[notice] To update, run: python3.11 -m pip install --upgrade pip
Note: you may need to restart the kernel to use updated packages.

# Goal by Sterling
import matplotsoccer
import socceraction


a = df_shot

# Plot the actions
def nice_time(row):
    return f"{row.time_seconds}s"

a["nice_time"] = a.apply(nice_time, axis=1)
a["end_x"] = a["x"] + a["dx"]
a["end_y"] = a["y"] + a["dy"]
a["type"] = a.type_id.apply(lambda x: socceraction.atomic.spadl.config.actiontypes[x])
labels = a[["nice_time", "type", "player", "team"]]

ax = matplotsoccer.actions(
    location=a[["x", "y", "end_x", "end_y"]],
    action_type=a.type_id,
    team= a.team,
    # result= True,
    label=labels,
    labeltitle=["time", "actiontype", "player", "team"],
    zoom=False,
    figsize=10
)

Passes

#players names
df_player_name = df_1.groupby(["team", 'player_id', 'player']).id.count().reset_index()
df_player_name = df_player_name.set_index('player_id')
print(df_player_name)

               team            player   id
player_id                                 
28550.0     Chelsea      Thiago Silva   67
97692.0     Chelsea   Raheem Sterling   74
300426.0    Chelsea       Axel Disasi   88
361330.0    Chelsea       Reece James   56
362151.0    Chelsea    Robert Sánchez   60
363496.0    Chelsea    Marc Cucurella   68
369430.0    Chelsea    Enzo Fernández   59
375621.0    Chelsea   Conor Gallagher   73
393407.0    Chelsea     Armando Broja    6
395692.0    Chelsea       Cole Palmer   60
403850.0    Chelsea        Malo Gusto   34
404616.0    Chelsea  Lesley Ugochukwu    1
409376.0    Chelsea   Mykhailo Mudryk   10
410175.0    Chelsea    Moisés Caicedo   51
426050.0    Chelsea   Nicolas Jackson   28
69778.0    Man City       Kyle Walker   82
93894.0    Man City     Mateo Kovacic   19
113069.0   Man City     Jack Grealish   36
121774.0   Man City           Ederson   59
136741.0   Man City    Bernardo Silva   77
297390.0   Man City     Manuel Akanji   78
303139.0   Man City             Rodri  107
313171.0   Man City        Rúben Dias  108
315227.0   Man City    Erling Haaland   27
331254.0   Man City        Phil Foden   70
365409.0   Man City    Julián Álvarez   54
388098.0   Man City       Jérémy Doku   39
402664.0   Man City    Josko Gvardiol   90

ManCity passing network

[the full match with substitutions]

# define a passer and a pass recipient and add the columns
df_MC_pn_ext = df_1[df_1['team'] == 'Man City']
df_MC_pn_ext['passer'] = df_MC_pn_ext['player_id']
df_MC_pn_ext['recipient'] = df_MC_pn_ext['player_id'].shift(-1)

# define successful passes
MC_successful_passes_ext = df_MC_pn_ext[(df_MC_pn_ext['type'] == 'Pass')&(df_MC_pn_ext['outcome_type'] == 'Successful')]
                                
    
# define an average player's location and quantity of passes 
MC_average_locations_ext = MC_successful_passes_ext.groupby('passer').agg({'x': ['mean'], 'y': ['mean','count']})

MC_average_locations_ext.columns = ['x', 'y', 'count']

MC_average_locations_ext= MC_average_locations_ext.merge(df_player_name[['player']], how='left', left_on='passer', right_index=True) \
        .rename(columns={'player': 'player_name'}) 

# count passes
pass_between_MC_ext = MC_successful_passes_ext.groupby(['passer', 'recipient']).id.count().reset_index()
pass_between_MC_ext.rename({'id':'pass_count'}, axis='columns', inplace=True)

pass_between_MC_ext = pass_between_MC_ext.merge(MC_average_locations_ext, left_on ='passer', right_index = True)
pass_between_MC_ext = pass_between_MC_ext.merge(MC_average_locations_ext, left_on ='recipient', right_index = True, suffixes=['', '_end'])


print(pass_between_MC_ext)

#Put the pass network on the pitch
pitch = Pitch(pitch_type='opta', positional=True, shade_middle=True, positional_color='#eadddd', shade_color='#f2f2f2', axis=True, label=True, tick=True, line_zorder =2 )  
fig, ax = plt.subplots(figsize=(13, 8.5))

pitch.draw(ax=ax)


arrows = pitch.arrows(pass_between_MC_ext.x, pass_between_MC_ext.y, pass_between_MC_ext.x_end, pass_between_MC_ext.y_end, ax=ax, \
                      width =3, headwidth=3, color = 'red', zorder = 1, alpha = 0.5)

nodes = pitch.scatter(MC_average_locations_ext.x, MC_average_locations_ext.y, s =100, color = "blue", linewidth = 2.5, \
                      alpha =1, zorder = 1, ax=ax)

for i, row in MC_average_locations_ext.iterrows():
    pitch.annotate(row.player_name, (MC_average_locations_ext.x[i], MC_average_locations_ext.y[i]), c='black', \
                   weight = "bold", size=10, ax=ax, zorder = 4)

/var/folders/sc/p5r88fzj46n6sr9l_c6dgt540000gn/T/ipykernel_43692/1285356849.py:3: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df_MC_pn_ext['passer'] = df_MC_pn_ext['player_id']
/var/folders/sc/p5r88fzj46n6sr9l_c6dgt540000gn/T/ipykernel_43692/1285356849.py:4: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df_MC_pn_ext['recipient'] = df_MC_pn_ext['player_id'].shift(-1)

       passer  recipient  pass_count          x          y  count  \
0     69778.0   121774.0           6  53.659091  16.763636     44   
15   113069.0   121774.0           1  71.540000  79.460000     15   
34   136741.0   121774.0           2  55.897917  62.527083     48   
46   297390.0   121774.0           5  41.710526  67.431579     57   
57   303139.0   121774.0           3  52.832836  46.425373     67   
..        ...        ...         ...        ...        ...    ...   
64   303139.0   388098.0           4  52.832836  46.425373     67   
75   313171.0   388098.0           2  40.831579  37.307895     76   
83   315227.0   388098.0           1  61.230000  52.780000     10   
101  365409.0   388098.0           2  64.851515  32.566667     33   
119  402664.0   388098.0           6  52.310000  84.065000     60   

        player_name  x_end  y_end  count_end player_name_end  
0       Kyle Walker  14.69  54.69         40         Ederson  
15    Jack Grealish  14.69  54.69         40         Ederson  
34   Bernardo Silva  14.69  54.69         40         Ederson  
46    Manuel Akanji  14.69  54.69         40         Ederson  
57            Rodri  14.69  54.69         40         Ederson  
..              ...    ...    ...        ...             ...  
64            Rodri  68.44  70.83         10     Jérémy Doku  
75       Rúben Dias  68.44  70.83         10     Jérémy Doku  
83   Erling Haaland  68.44  70.83         10     Jérémy Doku  
101  Julián Álvarez  68.44  70.83         10     Jérémy Doku  
119  Josko Gvardiol  68.44  70.83         10     Jérémy Doku  

[120 rows x 11 columns]

Chelsea passing network

[the full match with substitutions]

# define a passer and a pass recipient and add the columns
df_Ch_pn_ext = df_1[df_1['team'] == 'Chelsea']
df_Ch_pn_ext['passer'] = df_Ch_pn_ext['player_id']
df_Ch_pn_ext['recipient'] = df_Ch_pn_ext['player_id'].shift(-1)

# define successful passes
Ch_successful_passes_ext = df_Ch_pn_ext[(df_Ch_pn_ext['type'] == 'Pass')&(df_Ch_pn_ext['outcome_type'] == 'Successful')]

# define an average player's location and quantity of passes 
Ch_average_locations_ext = Ch_successful_passes_ext.groupby('passer').agg({'x': ['mean'], 'y': ['mean','count']})
Ch_average_locations_ext.columns = ['x', 'y', 'count']

Ch_average_locations_ext= Ch_average_locations_ext.merge(df_player_name[['player']], how='left', left_on='passer', right_index=True) \
        .rename(columns={'player': 'player_name'}) 

# count passes
pass_between_Ch_ext = Ch_successful_passes_ext.groupby(['passer', 'recipient']).id.count().reset_index()
pass_between_Ch_ext.rename({'id':'pass_count'}, axis='columns', inplace=True)

pass_between_Ch_ext = pass_between_Ch_ext.merge(Ch_average_locations_ext, left_on ='passer', right_index = True)
pass_between_Ch_ext = pass_between_Ch_ext.merge(Ch_average_locations_ext, left_on ='recipient', right_index = True, suffixes=['', '_end'])
print(pass_between_Ch_ext)

# put the pass network on the pitch
pitch = Pitch(pitch_type='opta', positional=True, shade_middle=True, positional_color='#eadddd', shade_color='#f2f2f2', axis=True, label=True, tick=True, line_zorder =2 )  
fig, ax = plt.subplots(figsize=(13, 8.5))

pitch.draw(ax=ax)

arrows = pitch.arrows(pass_between_Ch_ext.x, pass_between_Ch_ext.y, pass_between_Ch_ext.x_end, pass_between_Ch_ext.y_end, ax=ax, width =3, headwidth=3, color = 'blue', zorder = 1, alpha = 0.5)

nodes = pitch.scatter(Ch_average_locations_ext.x, Ch_average_locations_ext.y, s =300, color = "grey", linewidth = 2.5, alpha =1, zorder = 1, ax=ax)

for i, row in Ch_average_locations_ext.iterrows():
    pitch.annotate(row.player_name, (Ch_average_locations_ext.x[i], Ch_average_locations_ext.y[i]), c='black', \
                   weight = "bold", size=10, ax=ax, zorder = 4)

/var/folders/sc/p5r88fzj46n6sr9l_c6dgt540000gn/T/ipykernel_43692/4097516628.py:3: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df_Ch_pn_ext['passer'] = df_Ch_pn_ext['player_id']
/var/folders/sc/p5r88fzj46n6sr9l_c6dgt540000gn/T/ipykernel_43692/4097516628.py:4: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  df_Ch_pn_ext['recipient'] = df_Ch_pn_ext['player_id'].shift(-1)

      passer  recipient  pass_count          x          y  count  \
0    28550.0    28550.0           1  28.574510  63.405882     51   
17  300426.0    28550.0          15  26.877273  30.542424     66   
28  361330.0    28550.0           2  43.459375  14.106250     32   
36  362151.0    28550.0           6   9.584375  46.896875     32   
44  363496.0    28550.0           8  37.200000  87.405405     37   
..       ...        ...         ...        ...        ...    ...   
74  395692.0    97692.0           4  50.322222  35.629630     27   
83  403850.0    97692.0           4  41.656000  12.572000     25   
91  410175.0    97692.0           1  39.543333  46.743333     30   
50  363496.0   409376.0           2  37.200000  87.405405     37   
71  375621.0   409376.0           1  52.740000  46.137500     40   

        player_name      x_end      y_end  count_end  player_name_end  
0      Thiago Silva  28.574510  63.405882         51     Thiago Silva  
17      Axel Disasi  28.574510  63.405882         51     Thiago Silva  
28      Reece James  28.574510  63.405882         51     Thiago Silva  
36   Robert Sánchez  28.574510  63.405882         51     Thiago Silva  
44   Marc Cucurella  28.574510  63.405882         51     Thiago Silva  
..              ...        ...        ...        ...              ...  
74      Cole Palmer  63.793333  58.693333         15  Raheem Sterling  
83       Malo Gusto  63.793333  58.693333         15  Raheem Sterling  
91   Moisés Caicedo  63.793333  58.693333         15  Raheem Sterling  
50   Marc Cucurella  65.850000  86.600000          2  Mykhailo Mudryk  
71  Conor Gallagher  65.850000  86.600000          2  Mykhailo Mudryk  

[101 rows x 11 columns]

# put the pass network on the pitch
pitch = Pitch(pitch_type='opta', positional=True, shade_middle=True, positional_color='#eadddd', shade_color='#f2f2f2', axis=True, label=True, tick=True, line_zorder =2 )  
fig, ax = plt.subplots(figsize=(13, 8.5))

pitch.draw(ax=ax)

arrows = pitch.arrows(pass_between_Ch_ext.x, pass_between_Ch_ext.y, pass_between_Ch_ext.x_end, pass_between_Ch_ext.y_end, ax=ax, width =3, headwidth=3, color = 'blue', zorder = 1, alpha = 0.5)

nodes = pitch.scatter(Ch_average_locations_ext.x, Ch_average_locations_ext.y, s =300, color = "grey", linewidth = 2.5, alpha =1, zorder = 1, ax=ax)

for i, row in Ch_average_locations_ext.iterrows():
    pitch.annotate(row.player_name, (Ch_average_locations_ext.x[i], Ch_average_locations_ext.y[i]), c='black', \
                   weight = "bold", size=10, ax=ax, zorder = 4)

Passes by players

Chelsea players passes

# prepare the dataframe of passes by Chelsea
mask_chelsea = (df_1.type == 'Pass') & (df_1.team == "Chelsea")
df_passes_ch = df_1.loc[mask_chelsea, ['x', 'y', 'end_x', 'end_y', 'player', 'outcome_type']]
#get the list of all players who made a pass
names_ch = df_passes_ch['player'].unique()

# draw 4x4 pitches
pitch = Pitch(pitch_type='opta', line_color='black', pad_top=20)
fig, axs = pitch.grid(ncols = 4, nrows = 4, figheight=12,  grid_height=0.85, title_height=0.06, axis=False,
                     endnote_height=0.04, title_space=0.04, endnote_space=0)

# for each player
for name, ax in zip(names_ch, axs['pitch'].flat[:len(names_ch)]):
    #put player name over the plot
    ax.text(60, -10, name,
            ha='center', va='center', fontsize=14)
    #take only successful passes by this player
    player_df_sc_ch = df_passes_ch.loc[(df_passes_ch["player"] == name) & (df_passes_ch["outcome_type"] == "Successful")]
    #take only unsuccessful passes by this player
    player_df_unsc_ch = df_passes_ch.loc[(df_passes_ch["player"] == name) & (df_passes_ch["outcome_type"] == "Unsuccessful")]
    
    #scatter
    pitch.scatter(player_df_sc_ch.x, player_df_sc_ch.y, alpha = 0.2, s = 50, color = "green", ax=ax)
    pitch.scatter(player_df_unsc_ch.x, player_df_unsc_ch.y, alpha = 0.2, s = 50, color = "red", ax=ax)
    #plot arrow
    pitch.arrows(player_df_sc_ch.x, player_df_sc_ch.y,
            player_df_sc_ch.end_x, player_df_sc_ch.end_y, color = "green", ax=ax, width=1)
    pitch.arrows(player_df_unsc_ch.x, player_df_unsc_ch.y,
            player_df_unsc_ch.end_x, player_df_unsc_ch.end_y, color = "red", ax=ax, width=1)

# We have more than enough pitches - remove them
for ax in axs['pitch'][-1, 16 - len(names_ch)-2:]:
    ax.remove()

    
# Another way to set title using mplsoccer
axs['title'].text(0.5, 0.5, 'Chelsea passes against Man City', ha='center', va='center', fontsize=30)
plt.show()

ManCity players passes

# prepare the dataframe of passes by Man City
mask_mc = (df_1.type == 'Pass') & (df_1.team == "Man City") 
df_passes_mc = df_1.loc[mask_mc, ['x', 'y', 'end_x', 'end_y', 'player', "outcome_type"]]
# get the list of all players who made a pass
names_mc = df_passes_mc['player'].unique()

# draw 4x4 pitches
pitch = Pitch(pitch_type='opta', line_color='black', pad_top=20)
fig, axs = pitch.grid(ncols = 4, nrows = 4, figheight=12,  grid_height=0.85, title_height=0.06, axis=False,
                     endnote_height=0.04, title_space=0.04, endnote_space=0)

# for each player
for name, ax in zip(names_mc, axs['pitch'].flat[:len(names_mc)]):
    # put player name over the plot
    ax.text(60, -10, name,
            ha='center', va='center', fontsize=14)
    #take only successful passes by this player
    player_df_sc_mc = df_passes_mc.loc[(df_passes_mc["player"] == name) & (df_passes_mc["outcome_type"] == "Successful")]
    #take only unsuccessful passes by this player
    player_df_unsc_mc = df_passes_mc.loc[(df_passes_mc["player"] == name) & (df_passes_mc["outcome_type"] == "Unsuccessful")]
    
    #scatter
    pitch.scatter(player_df_sc_mc.x, player_df_sc_mc.y, alpha = 0.2, s = 50, color = "green", ax=ax)
    pitch.scatter(player_df_unsc_mc.x, player_df_unsc_mc.y, alpha = 0.2, s = 50, color = "red", ax=ax)
    #plot arrow
    pitch.arrows(player_df_sc_mc.x, player_df_sc_mc.y,
            player_df_sc_mc.end_x, player_df_sc_mc.end_y, color = "green", ax=ax, width=1)
    pitch.arrows(player_df_unsc_mc.x, player_df_unsc_mc.y,
            player_df_unsc_mc.end_x, player_df_unsc_mc.end_y, color = "red", ax=ax, width=1)

# We have more than enough pitches - remove them
for ax in axs['pitch'][-1, 16 - len(names_mc)-2:]:
    ax.remove()

    
# Another way to set title using mplsoccer
axs['title'].text(0.5, 0.5, 'Man City passes against Chelsea', ha='center', va='center', fontsize=30)
plt.show()

df_passes_mc_ratio = df_passes_mc.reset_index().groupby(['player', 'outcome_type']).index.count()

df_passes_mc_ratio

player          outcome_type
Bernardo Silva  Successful      48
                Unsuccessful     9
Ederson         Successful      40
                Unsuccessful     3
Erling Haaland  Successful      10
                Unsuccessful     2
Jack Grealish   Successful      15
                Unsuccessful     3
Josko Gvardiol  Successful      60
                Unsuccessful     3
Julián Álvarez  Successful      33
                Unsuccessful     9
Jérémy Doku     Successful      10
                Unsuccessful     5
Kyle Walker     Successful      44
                Unsuccessful    11
Manuel Akanji   Successful      57
                Unsuccessful     4
Mateo Kovacic   Successful      11
Phil Foden      Successful      33
                Unsuccessful     3
Rodri           Successful      67
                Unsuccessful     9
Rúben Dias      Successful      76
                Unsuccessful     8
Name: index, dtype: int64

Passes matrix

def get_pass_matrix():
    pass_init = events_atomic_spadl[events_atomic_spadl["type_id"] == 0][
        ["type_id", "player", "original_event_id", "team_id"]
    ]
    pass_receive = events_atomic_spadl[events_atomic_spadl["type_id"] == 23][
        ["type_id", "player", "original_event_id"]
    ]
    # return pass_init
    pass_matrix = pass_init.join(
        pass_receive.set_index("original_event_id"),
        on="original_event_id",
        rsuffix="_other",
    )[["player", "player_other", "team_id"]]
    pass_matrix["count"] = 1
    pass_matrix = pass_matrix.groupby(["team_id", "player", "player_other"]).agg(
        {"count": "sum"}
    )
    return pass_matrix.reset_index()


pass_matrix = get_pass_matrix()

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt


def print_pass_matrix(df, team):
    # Create a pivot table to reshape the data for the pass matrix
    pass_matrix = df.pivot_table(
        index="player", columns="player_other", values="count", fill_value=0
    )

    # Get players and pass counts
    players = pass_matrix.columns
    pass_counts = pass_matrix.values

    # Plotting the pass matrix
    fig, ax = plt.subplots()
    im = ax.imshow(pass_counts, cmap="Blues")

    # Set ticks and labels
    ax.set_xticks(np.arange(len(players)))
    ax.set_yticks(np.arange(len(players)))
    ax.set_xticklabels(players)
    ax.set_yticklabels(players)

    # Rotate the tick labels and set their alignment
    plt.setp(ax.get_xticklabels(), rotation=45, ha="right", rotation_mode="anchor")

    # Loop over data dimensions and create text annotations
    for i in range(len(players)):
        for j in range(len(players)):
            text = ax.text(
                j, i, pass_counts[i, j], ha="center", va="center", color="black"
            )

    # Set title and show the plot
    ax.set_title(f"{team} Pass Matrix")
    fig.tight_layout()
    plt.show()

print_pass_matrix(pass_matrix[pass_matrix['team_id'] == MC_team_id], 'City')

print_pass_matrix(pass_matrix[pass_matrix['team_id'] == chelsea_team_id], 'Chelsea')

Heatmap

# define dataframes by the teams
df_MC_hm = df_1[(df_1['team'] == 'Man City')&(df_1['outcome_type'] == 'Successful')][['minute', 'second', 'team', 'player','type', 'outcome_type','x','y', 'end_x','end_y']]
df_Ch_hm = df_1[(df_1['team'] == 'Chelsea')&(df_1['outcome_type'] == 'Successful')][['minute', 'second', 'team', 'player','type', 'outcome_type','x','y', 'end_x','end_y']]

Heatmap for ManCity

# make a pitch with data from OPTA and with a mention of the thirds and draw the heatmap for MC
pitch = Pitch(pitch_type='opta', positional=True, shade_middle=True, positional_color='#eadddd', shade_color='#f2f2f2', axis=True, label=True, tick=True, line_zorder =2 )  
fig, ax = pitch.draw()
pitch.kdeplot(df_MC_hm['x'], df_MC_hm['y'], ax=ax, cmap = 'Reds', zorder=1)

<Axes: >

Heatmap for Chelsea

# make a pitch with data from OPTA and with a mention of the thirds and draw the heatmap for Chelsea
pitch = Pitch(pitch_type='opta', positional=True, shade_middle=True, positional_color='#eadddd', shade_color='#f2f2f2', axis=True, label=True, tick=True, line_zorder =2 )  
fig, ax = pitch.draw()
pitch.kdeplot(100 - df_Ch_hm['x'], 100 - df_Ch_hm['y'], ax=ax, cmap = 'Blues', zorder=1)

<Axes: >

xG Flow Chat

import matplotlib as mpl
import matplotlib.pyplot as plt

df_2 = pd.read_csv('data/12112023/xG_shots.csv')

df_2

	period	minute	second	type	outcome_type	team_id	team	player_id	player	x	y	xG	goal (y/n)
0	FirstHalf	0	25	SavedShot	Successful	15	Chelsea	361330	Reece James	85.9	28.8	0.03	n
1	FirstHalf	11	37	SavedShot	Successful	15	Chelsea	395692	Cole Palmer	87.6	31.4	0.04	n
2	FirstHalf	15	35	SavedShot	Successful	167	Man City	365409	Juli?n ?lvarez	93.8	29.8	0.04	n
3	FirstHalf	16	6	SavedShot	Successful	167	Man City	402664	Josko Gvardiol	86.7	57.8	0.01	n
4	FirstHalf	18	14	SavedShot	Successful	15	Chelsea	375621	Conor Gallagher	77.7	44.2	0.05	n
5	FirstHalf	24	3	Goal	Successful	167	Man City	315227	Erling Haaland	88.5	50.0	0.79	y
6	FirstHalf	26	26	SavedShot	Successful	15	Chelsea	410175	Mois?s Caicedo	77.8	53.3	0.03	n
7	FirstHalf	27	50	SavedShot	Successful	15	Chelsea	361330	Reece James	77.5	64.9	0.06	n
8	FirstHalf	28	24	Goal	Successful	15	Chelsea	28550	Thiago Silva	95.2	39.1	0.07	y
9	FirstHalf	31	40	MissedShot	Successful	167	Man City	315227	Erling Haaland	99.2	58.4	0.07	n
10	FirstHalf	33	51	MissedShot	Successful	167	Man City	331254	Phil Foden	85.6	36.1	0.07	n
11	FirstHalf	36	23	Goal	Successful	15	Chelsea	97692	Raheem Sterling	96.7	50.9	0.43	y
12	FirstHalf	39	48	SavedShot	Successful	15	Chelsea	369430	Enzo Fern?ndez	78.5	63.2	0.04	n
13	FirstHalf	40	41	MissedShot	Successful	15	Chelsea	97692	Raheem Sterling	88.3	69.3	0.02	n
14	FirstHalf	41	33	SavedShot	Successful	167	Man City	315227	Erling Haaland	87.4	55.8	0.31	n
15	FirstHalf	42	36	SavedShot	Successful	167	Man City	331254	Phil Foden	78.1	33.1	0.02	n
16	FirstHalf	43	14	SavedShot	Successful	15	Chelsea	426050	Nicolas Jackson	84.7	52.8	0.09	n
17	FirstHalf	45	27	Goal	Successful	167	Man City	297390	Manuel Akanji	93.0	51.2	0.29	y
18	FirstHalf	51	26	SavedShot	Successful	15	Chelsea	426050	Nicolas Jackson	89.9	66.3	0.04	n
19	SecondHalf	46	31	Goal	Successful	167	Man City	315227	Erling Haaland	97.7	52.8	0.87	y
20	SecondHalf	51	53	SavedShot	Successful	167	Man City	365409	Juli?n ?lvarez	86.3	51.5	0.12	n
21	SecondHalf	57	27	SavedShot	Successful	167	Man City	388098	Jeremy Doku	90.6	63.9	0.09	n
22	SecondHalf	59	42	SavedShot	Successful	15	Chelsea	395692	Cole Palmer	89.0	53.0	0.48	n
23	SecondHalf	60	50	SavedShot	Successful	167	Man City	331254	Phil Foden	85.4	38.6	0.07	n
24	SecondHalf	66	22	SavedShot	Successful	15	Chelsea	375621	Conor Gallagher	73.0	53.3	0.03	n
25	SecondHalf	66	25	Goal	Successful	15	Chelsea	426050	Nicolas Jackson	90.5	47.6	0.55	y
26	SecondHalf	74	53	MissedShot	Successful	15	Chelsea	403850	Malo Gusto	90.5	36.5	0.20	n
27	SecondHalf	80	17	MissedShot	Successful	15	Chelsea	375621	Conor Gallagher	71.0	47.9	0.02	n
28	SecondHalf	85	30	SavedShot	Successful	167	Man City	93894	Mateo Kovacic	80.6	49.5	0.03	n
29	SecondHalf	85	33	Goal	Successful	167	Man City	303139	Rodri	76.7	50.8	0.03	y
30	SecondHalf	94	24	Goal	Successful	15	Chelsea	395692	Cole Palmer	88.5	50.0	0.79	y
31	SecondHalf	98	1	MissedShot	Successful	167	Man City	69778	Kyle Walker	74.9	55.2	0.07	n

a_xG = [0]
h_xG = [0]
a_min = [0]
h_min = [0]

# Find home and away teams
hteam = df_2['team'].iloc[0]  # first row (Chelsea)
ateam = df_2['team'].iloc[-1] # last row (ManCity)

for x in range(len(df_2['xG'])):
    if df_2['team'][x]==ateam:
        a_xG.append(df_2['xG'][x])
        a_min.append(df_2['minute'][x])
    if df_2['team'][x]==hteam:
        h_xG.append(df_2['xG'][x])
        h_min.append(df_2['minute'][x])

a_xG

[0,
 0.04,
 0.01,
 0.79,
 0.07,
 0.07,
 0.31,
 0.02,
 0.29,
 0.87,
 0.12,
 0.09,
 0.07,
 0.03,
 0.03,
 0.07]

def nums_cumulative_sum(nums_list):
    return [sum(nums_list[:i+1]) for i in range(len(nums_list))]
a_cumulative = nums_cumulative_sum(a_xG)
h_cumulative = nums_cumulative_sum(h_xG)

#a_cumulative

#h_cumulative

a_total = round(a_cumulative [-1], 2)
h_total = round(h_cumulative [-1], 2)

# cumulative Sum xG home team
h_df = df_2[df_2['team'] == hteam]
h_df.sort_values(by='minute', inplace=True)
h_df['h_cum'] = h_df['xG'].cumsum()

# cumulative Sum xG away team
a_df = df_2[df_2['team'] == ateam]
a_df.sort_values(by='minute', inplace=True)
a_df['a_cum'] = a_df['xG'].cumsum()

/var/folders/sc/p5r88fzj46n6sr9l_c6dgt540000gn/T/ipykernel_43692/2695564966.py:3: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  h_df.sort_values(by='minute', inplace=True)
/var/folders/sc/p5r88fzj46n6sr9l_c6dgt540000gn/T/ipykernel_43692/2695564966.py:4: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  h_df['h_cum'] = h_df['xG'].cumsum()
/var/folders/sc/p5r88fzj46n6sr9l_c6dgt540000gn/T/ipykernel_43692/2695564966.py:8: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  a_df.sort_values(by='minute', inplace=True)
/var/folders/sc/p5r88fzj46n6sr9l_c6dgt540000gn/T/ipykernel_43692/2695564966.py:9: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  a_df['a_cum'] = a_df['xG'].cumsum()

h_df

	period	minute	second	type	outcome_type	team_id	team	player_id	player	x	y	xG	goal (y/n)	h_cum
0	FirstHalf	0	25	SavedShot	Successful	15	Chelsea	361330	Reece James	85.9	28.8	0.03	n	0.03
1	FirstHalf	11	37	SavedShot	Successful	15	Chelsea	395692	Cole Palmer	87.6	31.4	0.04	n	0.07
4	FirstHalf	18	14	SavedShot	Successful	15	Chelsea	375621	Conor Gallagher	77.7	44.2	0.05	n	0.12
6	FirstHalf	26	26	SavedShot	Successful	15	Chelsea	410175	Mois?s Caicedo	77.8	53.3	0.03	n	0.15
7	FirstHalf	27	50	SavedShot	Successful	15	Chelsea	361330	Reece James	77.5	64.9	0.06	n	0.21
8	FirstHalf	28	24	Goal	Successful	15	Chelsea	28550	Thiago Silva	95.2	39.1	0.07	y	0.28
11	FirstHalf	36	23	Goal	Successful	15	Chelsea	97692	Raheem Sterling	96.7	50.9	0.43	y	0.71
12	FirstHalf	39	48	SavedShot	Successful	15	Chelsea	369430	Enzo Fern?ndez	78.5	63.2	0.04	n	0.75
13	FirstHalf	40	41	MissedShot	Successful	15	Chelsea	97692	Raheem Sterling	88.3	69.3	0.02	n	0.77
16	FirstHalf	43	14	SavedShot	Successful	15	Chelsea	426050	Nicolas Jackson	84.7	52.8	0.09	n	0.86
18	FirstHalf	51	26	SavedShot	Successful	15	Chelsea	426050	Nicolas Jackson	89.9	66.3	0.04	n	0.90
22	SecondHalf	59	42	SavedShot	Successful	15	Chelsea	395692	Cole Palmer	89.0	53.0	0.48	n	1.38
24	SecondHalf	66	22	SavedShot	Successful	15	Chelsea	375621	Conor Gallagher	73.0	53.3	0.03	n	1.41
25	SecondHalf	66	25	Goal	Successful	15	Chelsea	426050	Nicolas Jackson	90.5	47.6	0.55	y	1.96
26	SecondHalf	74	53	MissedShot	Successful	15	Chelsea	403850	Malo Gusto	90.5	36.5	0.20	n	2.16
27	SecondHalf	80	17	MissedShot	Successful	15	Chelsea	375621	Conor Gallagher	71.0	47.9	0.02	n	2.18
30	SecondHalf	94	24	Goal	Successful	15	Chelsea	395692	Cole Palmer	88.5	50.0	0.79	y	2.97

h_goal = h_df[h_df['type'].str.contains("Goal")]
h_goal["scorechart"] = h_goal["minute"].astype(str) + "'" + " " +h_goal["player"]
a_goal = a_df[a_df['type'].str.contains("Goal")]
a_goal["scorechart"] = a_goal["minute"].astype(str) + "'" + " " +a_goal["player"]

/var/folders/sc/p5r88fzj46n6sr9l_c6dgt540000gn/T/ipykernel_43692/1230918128.py:2: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  h_goal["scorechart"] = h_goal["minute"].astype(str) + "'" + " " +h_goal["player"]
/var/folders/sc/p5r88fzj46n6sr9l_c6dgt540000gn/T/ipykernel_43692/1230918128.py:4: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  a_goal["scorechart"] = a_goal["minute"].astype(str) + "'" + " " +a_goal["player"]

h_goal

	period	minute	second	type	outcome_type	team_id	team	player_id	player	x	y	xG	goal (y/n)	h_cum	scorechart
8	FirstHalf	28	24	Goal	Successful	15	Chelsea	28550	Thiago Silva	95.2	39.1	0.07	y	0.28	28' Thiago Silva
11	FirstHalf	36	23	Goal	Successful	15	Chelsea	97692	Raheem Sterling	96.7	50.9	0.43	y	0.71	36' Raheem Sterling
25	SecondHalf	66	25	Goal	Successful	15	Chelsea	426050	Nicolas Jackson	90.5	47.6	0.55	y	1.96	66' Nicolas Jackson
30	SecondHalf	94	24	Goal	Successful	15	Chelsea	395692	Cole Palmer	88.5	50.0	0.79	y	2.97	94' Cole Palmer

xG flow without goals

fig, ax = plt.subplots(figsize = (10,5))
fig.set_facecolor('#3d4849')
ax.patch.set_facecolor('#3d4849')

plt.xticks([0,15,30,45,60,75,90])
plt.xlabel('Minute')
plt.ylabel('xG')

ax.step(x=a_min, y=a_cumulative, where='post')
ax.step(x=h_min, y=h_cumulative, where='post')


title_text = 'Chelsea-ManCity_xG flow'
ax.set_title(title_text, color='white')

ax.text(0.25, 0.98, 'Man City', color=(72/255, 147/255, 206/255), fontsize=10, ha='center', va='center', transform=ax.transAxes)
ax.text(0.75, 0.98, 'Chelsea', color='orange', fontsize=10, ha='center', va='center', transform=ax.transAxes)

Text(0.75, 0.98, 'Chelsea')

xG flow with goals

fig, ax = plt.subplots(figsize=(15,6))

# Create line plots
ax.step(x = h_df['minute'], y = h_df['h_cum'], where = 'post', color='orange')  
ax.step(x = a_df['minute'], y = a_df['a_cum'], where = 'post', color=(72/255, 147/255, 206/255))

# Create scatter plot for highlighting the goal
ax.scatter(x= h_goal['minute'], y = h_goal['h_cum'] , marker= 'o', s= 200, color='orange')
ax.scatter(x= a_goal['minute'], y = a_goal['a_cum'] , marker= 'o', s =200, color=(72/255, 147/255, 206/255))


# Customize our chart
for j, txt in h_goal['scorechart'].items():
    ax.annotate(txt, (h_goal['minute'][j], h_goal['h_cum'][j]), xycoords='data', ha='center',
                xytext=(-90, 50), textcoords='offset points',
                arrowprops=dict(arrowstyle="->", connectionstyle="arc,angleA=0,armA=50,rad=10", color='orange'))

for i, txt in a_goal['scorechart'].items():
    ax.annotate(txt, (a_goal['minute'][i], a_goal['a_cum'][i]), xycoords='data', ha='center',
                xytext=(-90, 50), textcoords='offset points',
                arrowprops=dict(arrowstyle="->", connectionstyle="arc,angleA=0,armA=50,rad=10", color=(72/255, 147/255, 206/255)))


plt.xticks([0,15,30,45,60,75,90])
plt.yticks([0, 0.5, 1, 1.5, 2, 2.5, 3])
plt.grid()

fig.text(s=hteam + " - " + ateam + " (" + '4' + " - " + '4' + ")", x=0.125, y=0.97, fontsize=18,fontweight="bold")
fig.text(s='Premier League 2023/2024', x=0.125, y=0.92, fontsize=12)
legend = ax.legend(title = 'Total Expected Goals(xG)' ,loc='best', shadow=True)
legend._legend_box.align = "left"

plt.ylabel("Expected Goals (xG)", fontsize = 12, labelpad = 20)
plt.xlabel("Minutes", fontsize = 12, labelpad = 20)
plt.show()

No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.

Simulating match result

%pip install statsmodels

Requirement already satisfied: statsmodels in /usr/local/lib/python3.11/site-packages (0.14.1)
Requirement already satisfied: numpy<2,>=1.18 in /usr/local/lib/python3.11/site-packages (from statsmodels) (1.25.0)
Requirement already satisfied: scipy!=1.9.2,>=1.4 in /usr/local/lib/python3.11/site-packages (from statsmodels) (1.11.4)
Requirement already satisfied: pandas!=2.1.0,>=1.0 in /usr/local/lib/python3.11/site-packages (from statsmodels) (1.5.3)
Requirement already satisfied: patsy>=0.5.4 in /usr/local/lib/python3.11/site-packages (from statsmodels) (0.5.6)
Requirement already satisfied: packaging>=21.3 in /usr/local/lib/python3.11/site-packages (from statsmodels) (23.2)
Requirement already satisfied: python-dateutil>=2.8.1 in /Users/ivn/Library/Python/3.11/lib/python/site-packages (from pandas!=2.1.0,>=1.0->statsmodels) (2.8.2)
Requirement already satisfied: pytz>=2020.1 in /usr/local/lib/python3.11/site-packages (from pandas!=2.1.0,>=1.0->statsmodels) (2023.3.post1)
Requirement already satisfied: six in /Users/ivn/Library/Python/3.11/lib/python/site-packages (from patsy>=0.5.4->statsmodels) (1.16.0)

[notice] A new release of pip is available: 23.3.1 -> 23.3.2
[notice] To update, run: python3.11 -m pip install --upgrade pip
Note: you may need to restart the kernel to use updated packages.

#https://soccermatics.readthedocs.io/en/latest/gallery/lesson5/plot_SimulateMatches.html
import statsmodels.api as sm
import statsmodels.formula.api as smf
#import pandas as pd
#import matplotlib.pyplot as plt
import numpy as np
import seaborn
from scipy.stats import poisson,skellam

# prediction before the match
epl = pd.read_csv('data/12112023/E0.csv')
epl = epl[['HomeTeam','AwayTeam','FTHG','FTAG']]
epl = epl.rename(columns={'FTHG': 'HomeGoals', 'FTAG': 'AwayGoals'})

goal_model_data = pd.concat([epl[['HomeTeam','AwayTeam','HomeGoals']].assign(home=1).rename(
            columns={'HomeTeam':'team', 'AwayTeam':'opponent','HomeGoals':'goals'}),
           epl[['AwayTeam','HomeTeam','AwayGoals']].assign(home=0).rename(
            columns={'AwayTeam':'team', 'HomeTeam':'opponent','AwayGoals':'goals'})])

poisson_model = smf.glm(formula="goals ~ home + team + opponent", data=goal_model_data,
                        family=sm.families.Poisson()).fit()
poisson_model.summary()

Generalized Linear Model Regression Results

Dep. Variable:	goals	No. Observations:	238
Model:	GLM	Df Residuals:	198
Model Family:	Poisson	Df Model:	39
Link Function:	Log	Scale:	1.0000
Method:	IRLS	Log-Likelihood:	-328.03
Date:	Tue, 30 Jan 2024	Deviance:	195.65
Time:	21:06:17	Pearson chi2:	166.
No. Iterations:	5	Pseudo R-squ. (CS):	0.3140
Covariance Type:	nonrobust

	coef	std err	z	P>|z|	[0.025	0.975]
Intercept	0.1748	0.389	0.450	0.653	-0.587	0.936
team[T.Aston Villa]	0.1510	0.279	0.541	0.588	-0.396	0.698
team[T.Bournemouth]	-0.7570	0.367	-2.061	0.039	-1.477	-0.037
team[T.Brentford]	-0.2635	0.305	-0.864	0.387	-0.861	0.334
team[T.Brighton]	-0.0214	0.288	-0.074	0.941	-0.585	0.542
team[T.Burnley]	-0.9117	0.392	-2.323	0.020	-1.681	-0.142
team[T.Chelsea]	-0.3181	0.321	-0.990	0.322	-0.948	0.312
team[T.Crystal Palace]	-0.7587	0.353	-2.147	0.032	-1.451	-0.066
team[T.Everton]	-0.6307	0.341	-1.849	0.064	-1.299	0.038
team[T.Fulham]	-0.8806	0.378	-2.327	0.020	-1.622	-0.139
team[T.Liverpool]	0.0633	0.285	0.222	0.825	-0.496	0.623
team[T.Luton]	-0.9074	0.378	-2.402	0.016	-1.648	-0.167
team[T.Man City]	0.1198	0.279	0.429	0.668	-0.427	0.667
team[T.Man United]	-0.7168	0.346	-2.074	0.038	-1.394	-0.039
team[T.Newcastle]	0.0462	0.285	0.162	0.871	-0.512	0.605
team[T.Nott'm Forest]	-0.5733	0.339	-1.692	0.091	-1.237	0.091
team[T.Sheffield United]	-0.8170	0.378	-2.163	0.031	-1.557	-0.077
team[T.Tottenham]	-0.0642	0.290	-0.221	0.825	-0.632	0.504
team[T.West Ham]	-0.1320	0.301	-0.438	0.661	-0.723	0.459
team[T.Wolves]	-0.4056	0.324	-1.250	0.211	-1.042	0.230
opponent[T.Aston Villa]	0.4374	0.405	1.080	0.280	-0.356	1.231
opponent[T.Bournemouth]	0.7051	0.377	1.873	0.061	-0.033	1.443
opponent[T.Brentford]	0.4586	0.402	1.142	0.254	-0.329	1.246
opponent[T.Brighton]	0.6030	0.391	1.544	0.123	-0.162	1.369
opponent[T.Burnley]	0.9058	0.370	2.450	0.014	0.181	1.630
opponent[T.Chelsea]	0.1347	0.436	0.309	0.758	-0.721	0.990
opponent[T.Crystal Palace]	0.3510	0.409	0.859	0.391	-0.450	1.152
opponent[T.Everton]	0.3699	0.409	0.904	0.366	-0.432	1.172
opponent[T.Fulham]	0.4674	0.394	1.186	0.236	-0.305	1.240
opponent[T.Liverpool]	-0.2018	0.453	-0.446	0.656	-1.089	0.685
opponent[T.Luton]	0.6155	0.389	1.584	0.113	-0.146	1.377
opponent[T.Man City]	-0.2087	0.481	-0.434	0.664	-1.151	0.734
opponent[T.Man United]	0.4075	0.410	0.994	0.320	-0.396	1.211
opponent[T.Newcastle]	0.0370	0.429	0.086	0.931	-0.803	0.877
opponent[T.Nott'm Forest]	0.3826	0.402	0.951	0.342	-0.406	1.171
opponent[T.Sheffield United]	0.9326	0.369	2.525	0.012	0.209	1.656
opponent[T.Tottenham]	0.4152	0.416	0.997	0.319	-0.401	1.231
opponent[T.West Ham]	0.5698	0.388	1.470	0.142	-0.190	1.330
opponent[T.Wolves]	0.5392	0.392	1.375	0.169	-0.229	1.308
home	0.2146	0.108	1.978	0.048	0.002	0.427

# set teams here
home_team='Chelsea'
away_team='Man City'

# predict for Chelsea vs. Manchester City
home_score_rate=poisson_model.predict(pd.DataFrame(data={'team': home_team, 'opponent': away_team,
                                       'home':1},index=[1]))
away_score_rate=poisson_model.predict(pd.DataFrame(data={'team': away_team, 'opponent': home_team,
                                       'home':0},index=[1]))
print(home_team + ' against ' + away_team + ' expect to score: ' + str(home_score_rate))
print(away_team + ' against ' + home_team + ' expect to score: ' + str(away_score_rate))

# lets just get a result
home_goals=np.random.poisson(home_score_rate)
away_goals=np.random.poisson(away_score_rate)
#print(home_team + ': ' + str(home_goals[0]))
#print(away_team + ': '  + str(away_goals[0]))

Chelsea against Man City expect to score: 1    0.871664
dtype: float64
Man City against Chelsea expect to score: 1    1.536197
dtype: float64

Expected result: Chelsea-ManCity 1-1